Fix UnicodeEncodeError issue using Python 2

sergioburdisso · Feb 19, 2020 · 867026e · 867026e
1 parent 8feeef5
commit 867026e
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 5 deletions.
diff --git a/pyss3/__init__.py b/pyss3/__init__.py
@@ -1772,6 +1772,8 @@ def learn(self, doc, cat, n_grams=1, prep=True, update=True):
         """
         try:
             doc = doc.decode(ENCODING)
+        except UnicodeEncodeError:  # for python 2 compatibility
+            doc = doc.encode(ENCODING).decode(ENCODING)
         except AttributeError:
             pass
         try:
@@ -1886,6 +1888,8 @@ def classify(self, doc, prep=True, sort=True, json=False):
 
         try:
             doc = doc.decode(ENCODING)
+        except UnicodeEncodeError:  # for python 2 compatibility
+            doc = doc.encode(ENCODING).decode(ENCODING)
         except BaseException:
             pass
 

diff --git a/tests/test_cmd_line.py b/tests/test_cmd_line.py
@@ -5,6 +5,8 @@
 from pyss3 import SS3
 from os import path
 
+import pyss3.cmd_line
+import pyss3.util
 import pytest
 import sys
 
@@ -22,11 +24,15 @@ def test_ss3prompt(mocker, monkeypatch):
     if PYTHON3:
         monkeypatch.setattr('builtins.input', lambda: 'Y')
     mocker.patch.object(SS3Prompt, "cmdloop")
-    mocker.patch("pyss3.cmd_line.STOPWORDS_FILE", "tests/ss3_models/ss3_stopwords[%s].txt")
-    mocker.patch(
-        "pyss3.util.EVAL_HTML_OUT_FILE",
-        "tests/ss3_models/ss3_model_evaluation[%s].html"
-    )
+    # not working in Python 2
+    # mocker.patch("pyss3.cmd_line.STOPWORDS_FILE", "tests/ss3_models/ss3_stopwords[%s].txt")
+    # mocker.patch(
+    #     "pyss3.util.EVAL_HTML_OUT_FILE",
+    #     "tests/ss3_models/ss3_model_evaluation[%s].html"
+    # )
+    # replaced by:
+    pyss3.cmd_line.STOPWORDS_FILE = "tests/ss3_models/ss3_stopwords[%s].txt"
+    pyss3.util.EVAL_HTML_OUT_FILE = "tests/ss3_models/ss3_model_evaluation[%s].html"
 
     main()